MPI and MCPL_input / ESS_butterfly_MCPL solution - duplicate stats "by file" if possible #432
Description
Suggestion from Konstantin Batkov during demo at ESS:
- Let the nodes read individual MCPL files if available - this allows increasing stats without biasing
TODO:
- ESS_butterfly_MCPL instrumentfile should check for existence of N independent files, otherwise issue warning and proceed by reading only the first file
- @EsbenKlinkby we should agree on a directory structure etc. for these files @ the ESS cluster
- Stats, weighting etc. handled according to the above - and should also depend on ssw2mcpl_app.c: Store NPS? mctools/mcpl#6